✨ Ability to add provider specific featureGates with Helm

[oprinmarius]

What this PR does / why we need it:
Ability to add provider specific featureGates.

Which issue(s) this PR fixes (optional, in fixes #<issue number>(, fixes #<issue_number>, ...) format, will close the issue(s) when PR gets merged):
Fixes #447

I used the following values:

image:
  manager:
    tag: v0.9.0
cert-manager:
  enabled: false
  installCRDs: false
addon: helm
infrastructure: "aws:v2.4.0"
core: "cluster-api:v1.6.2"
configSecret: 
  name: aws-variables
  namespace: default
manager:
  featureGates:
    core:
      ClusterTopology: true
      MachinePool: true     
    aws:
      EKSEnableIAM: true
      EKSAllowAddRoles: true
      ClusterTopology: true
      MachinePool: true

[netlify]

✅ Deploy Preview for kubernetes-sigs-cluster-api-operator ready!

Name	Link
🔨 Latest commit	ea3ba4c
🔍 Latest deploy log	https://app.netlify.com/sites/kubernetes-sigs-cluster-api-operator/deploys/6613d8aad4ce0e0008475543
😎 Deploy Preview	https://deploy-preview-448--kubernetes-sigs-cluster-api-operator.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify site configuration.

[zioproto]

How do you test this?

I tried:

cd cluster-api-operator/hack/charts
helm template --set infrastructure=azure:v1.13.2   \
                         --set addon=helm --set cert-manager.enabled=false \
                         --set image.manager.tag=v0.9.0 \
                         --set manager.featureGates.core.MachinePool=true  \
                         --set manager.featureGates.azure.MachinePool=true  \
                         --set manager.featureGates.azure.ClusterTopology=true \
                         --set manager.featureGates.azure.ClusterResourceSet=true \
                         --wait capi ./cluster-api-operator

But the template core.yaml is not rendered. However I don't pass additional --set values when I use the chart.

[oprinmarius]

How do you test this?

I tried:

cd cluster-api-operator/hack/charts
helm template --set infrastructure=azure:v1.13.2   \
                         --set addon=helm --set cert-manager.enabled=false \
                         --set image.manager.tag=v0.9.0 \
                         --set manager.featureGates.core.MachinePool=true  \
                         --set manager.featureGates.azure.MachinePool=true  \
                         --set manager.featureGates.azure.ClusterTopology=true \
                         --set manager.featureGates.azure.ClusterResourceSet=true \
                         --wait capi ./cluster-api-operator

But the template core.yaml is not rendered. However I don't pass additional --set values when I use the chart.

I used the following values:

image:
  manager:
    tag: v0.9.0
cert-manager:
  enabled: false
  installCRDs: false
addon: helm
infrastructure: "aws:v2.4.0"
core: "cluster-api:v1.6.2"
configSecret: 
  name: aws-variables
  namespace: default
manager:
  featureGates:
    core:
      ClusterTopology: true
      MachinePool: true     
    aws:
      EKSEnableIAM: true
      EKSAllowAddRoles: true
      ClusterTopology: true
      MachinePool: true

[oprinmarius]

@zioproto I think you need to specify core provider version or a secret for the template to be rendered:

cluster-api-operator/hack/charts/cluster-api-operator/templates/core.yaml

Line 40 in 006442f

{{- if or $coreVersion $.Values.configSecret.name }}

[alexander-demicev]

Hi, thanks for working on this. We have a set of e2e tests for ensuring helm chart works as expected https://github.com/kubernetes-sigs/cluster-api-operator/blob/main/test/e2e/helm_test.go. Can you add some tests for feature gates there?

[zioproto]

@alexander-demicev could you please review the implemented test ?

What I want you to double check is this comparison:

Expect(manifests).To(Equal(string(expectedManifests)))

It seems manifests generated with the helmChart.Run() function, is missing the capi-cluster-api-operator Deployment. This Deployment is actually there when using helm from command line.

For this reason I had to drop the deployment from expectedManifests:

35e6d85

I don't understand why the Deployment capi-cluster-api-operator is missing, please clarify if what is implemented here is correct. Thanks

[alexander-demicev]

@zioproto by default helmChart.Run(...) will only print resources that helm hooks, that's why only providers are in the output.

[alexander-demicev]

/approve

[jackfrancis]

Apologies for the blast

@Danil-Grigorev @Fedosin @alexander-demicev @damdo @furkatgofurov7 could we get an lgtm to land this fix and unblock some provider workflows that use cluster-api-operator?

Thank you!

[Danil-Grigorev]

/approve

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: alexander-demicev, Danil-Grigorev

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [Danil-Grigorev,alexander-demicev]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[Danil-Grigorev]

/lgtm

[k8s-ci-robot]

LGTM label has been added.

Git tree hash: 25d10c2517dbaa24ae69dcb1153bdc6688cdd466

[jackfrancis]

/cherrypick release-0.9

[k8s-infra-cherrypick-robot]

@jackfrancis: new pull request created: #484

In response to this:

/cherrypick release-0.9

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.